224 DEVELOPMENT AND HEREDITV. 



This is apparently what has happened in the phy- 

 logenetic development of Salpa. Here the stolon 

 is not only much diminished in size, but to prevent 

 its retarding the progress of the animal through 

 the water, it is coiled up and enveloped in the 

 wall of the body. When a portion of the stolon 

 has developed into nearly perfect individuals, it 

 becomes pushed out and hangs free in the water, 

 and is finally broken off. The first individual of 

 Salpa is therefore homologous with the first indi- 

 vidual of Perophera. The free-swimming life of 

 Salpa is most favourable to a single untrammelled 

 individual, or to a perfect colony ; accordingly these 

 two stages in the growth of the colony become 

 more pronounced, and the intermediate stages be- 

 come abbreviated and compressed. Instead of two 

 distinct generations of Salpa, we have only two 

 peculiarly modified forms of the same colony, which 

 eventually separate by force of circumstance. 



An extremely interesting example of meta- 

 morphosis is presented to us by the Medusas. The 

 study of morphology has shown us that these free- 

 swimming animals were originally derived from 

 stationary Hydroids. As was explained in the case 

 of the butterfly, this phylogenetic development 

 must have been gradual, and individual develop- 

 ment must have been originally without sudden 

 jumps or metamorphoses. The individual Medusa 



